If your Google Ads performance changed during August 2026 even though you barely touched the campaign, the cause might not be something you did.
Google began rolling out a change to its bidding systems on 17 August 2026. It affects certain campaigns that are Limited by budget and use target-based Smart Bidding strategies such as Target CPA or Target ROAS.
Google has warned that advertisers may see temporary changes in traffic and performance while the new behaviour rolls out.
That does not mean every poor week in August can be blamed on Google. But if an affected campaign suddenly started spending differently, delivering a higher cost per acquisition or moving closer to its stated ROAS target, this update is one of the first things worth checking.
The important question is not simply whether the numbers changed.
It is whether the campaign is now behaving more closely to the target you actually told Google to pursue.
What did Google change on 17 August 2026?
Google says it has updated the bidding behaviour of campaigns that are constrained by budget and use a target-based bidding strategy.
For Search campaigns, the most relevant strategies are:
- Target CPA;
- Target ROAS.
The wider rollout also affects eligible Shopping, Performance Max, Demand Gen and Travel campaigns. Google says Target CPC is included for Demand Gen campaigns.
Crucially, Google says this is not a change to the Google Ads auction itself. The auction has not suddenly been redesigned. The change is within the automated bidding system and how budget-limited campaigns work towards the targets advertisers have set.
Google’s official FAQ explains the August 2026 bidding changes in more detail.
Which Google Ads campaigns are actually affected?
Before assuming this update explains a change in performance, check whether your campaign fits the conditions.
The campaign needs to be:
- Limited by budget; and
- using an affected target-based bidding strategy such as Target CPA or Target ROAS.
If your Search campaign uses Target CPA or Target ROAS but is not constrained by budget, Google says its bidding behaviour is not changing as part of this update.
That distinction matters.
A drop in conversions on an unrestricted campaign could still have many explanations: demand, competition, tracking, conversion delays, landing-page changes or normal variation. This particular August update should not become a convenient explanation for every performance problem.
Google describes a campaign as Limited by budget when its average daily budget is restricting its ability to capture the traffic available under its current settings. In practice, the campaign could potentially show more often or generate more activity if additional budget were available.
If you are already working with a small budget, our article on whether £10 a day is enough for Google Ads explains why the relationship between budget, demand and campaign objectives matters.
Why might your CPA suddenly get worse?
This is where the update becomes more interesting.
Some budget-limited campaigns have historically performed substantially better than the Target CPA or Target ROAS entered into the account.
Google uses a simple Target CPA example.
Imagine you tell Google that you are prepared to pay £10 per conversion.
Despite that target, the campaign has recently been generating conversions at approximately £5 each.
That looks great.
But the £10 target is still telling Google’s bidding system that a conversion costing around £10 is acceptable.
Under the updated behaviour, a budget-limited campaign may begin operating more consistently towards the target that is actually configured. As a result, the real CPA could move upwards towards £10.
Google provides guidance on how advertisers should review their targets following the change.
This creates a slightly counterintuitive situation.
Your CPA can become worse on paper without anything necessarily being broken.
If the campaign moves from £5 to £8 per conversion while your stated target is £10, Google’s system may simply be making greater use of the flexibility you had already given it.
Your target might have been too loose all along
This is probably the most important lesson from the change.
A Target CPA or Target ROAS should represent something commercially meaningful.
It should not simply be a number that was entered months ago and forgotten.
If a business can profitably acquire a customer at £50, then a £50 Target CPA may be sensible.
If that same business actually needs acquisitions below £25 to remain profitable but has a £50 target configured, it has effectively told Google that much more expensive conversions are acceptable.
Historically outperforming that target can hide the problem.
If the bidding system now begins using more of the flexibility within the target, the mismatch suddenly becomes much more obvious.
The same principle applies to Target ROAS.
If the ROAS target in Google Ads does not reflect the return the business genuinely requires, tighter adherence to that target can expose a commercial problem that was already present in the account.
That is why we would not judge this update simply by asking whether CPA increased or ROAS decreased.
The better question is:
Was the target in the account actually the target the business wanted Google to pursue?
Should you immediately change your Target CPA or Target ROAS?
Not automatically.
Google has introduced a Target Adjustment Tool for eligible campaigns, allowing advertisers to review targets that may no longer reflect the level of performance they want to maintain.
If a campaign has consistently been outperforming its target and that historic efficiency is important to the business, adjusting the target closer to actual recent performance may make sense.
Using the earlier example:
- Target CPA: £10;
- recent actual CPA: £5;
- desired future CPA: approximately £5.
An advertiser might decide that the Target CPA should be moved closer to £5.
But that does not mean £5 is automatically the right answer either.
Perhaps £7 would still produce profitable customers while allowing the campaign to reach more potential customers. Perhaps £5 is essential because lead quality is inconsistent. Perhaps the business should be evaluating qualified leads or sales rather than form submissions in the first place.
The target should be based on business economics, not simply whichever historical Google Ads number looks nicest.
Should you just increase the budget instead?
Possibly, but again, not automatically.
A campaign being Limited by budget tells you there is more potential traffic available under the existing settings. It does not tell you that buying all of that additional traffic would be profitable.
Before increasing the budget, look at what the campaign is actually producing.
Are the conversions good leads?
Are those leads becoming customers?
Is the revenue or profit generated worth the acquisition cost?
Does the business have the capacity to handle more enquiries or orders?
If the campaign is producing genuinely profitable business at an acceptable acquisition cost, increasing budget may be entirely reasonable.
If the campaign is generating poor-quality leads, increasing the budget simply gives it more money to generate them.
Budget constraints are therefore only part of the diagnosis.
What shouldn’t you change just because performance moved?
A sudden movement in Google Ads performance creates a strong temptation to start editing everything.
Budget goes up.
Target CPA comes down.
Keywords get paused.
Bid controls are added.
Somebody changes the landing page.
Three days later, nobody knows whether the original performance change came from Google, normal volatility or the five additional changes made afterwards.
Google specifically advises advertisers not to apply data exclusions or new bid limits solely because of this rollout. Unnecessary bidding interventions can introduce additional performance fluctuations and make the account harder to evaluate.
That does not mean you should never make changes.
It means you should establish what actually happened first.
How long should you wait before judging the change?
Google recommends evaluating performance over one to two conversion cycles after changing a target or after the new bidding behaviour has taken effect.
That is more useful than applying an arbitrary rule such as “wait seven days”.
A conversion cycle is the amount of time it typically takes for a click to result in a recorded conversion.
For one business, most conversions might happen on the same day.
For another, somebody may click an advert, return several times and enquire a week later.
If your normal conversion cycle is long, evaluating Smart Bidding after two or three days can give a misleading picture because many of the eventual conversions have not happened yet.
Google’s automated bidding also needs data from the changed environment. Constantly changing settings while trying to assess the result makes that assessment harder.
Be careful with Google Ads forecasts during the rollout
There is another particularly important caveat for advertisers reviewing campaigns right now.
Google says its planning and forecasting systems are being updated to account for the new bidding behaviour.
During the transition period from 17 August to 31 August 2026, Google has warned that forecasts may contain inaccuracies.
That means estimates from planning tools should be treated with additional caution during this period.
If a forecast suddenly suggests a substantial change in expected conversions, CPA or spend, it is worth remembering that Google itself has acknowledged that the transition may temporarily affect forecast accuracy.
A higher CPA does not automatically mean the campaign became worse
This is an easy trap to fall into.
Imagine a campaign with:
- a Target CPA of £50;
- a historical actual CPA of £30.
After the new bidding behaviour takes effect, actual CPA moves to £42.
On the surface, performance has deteriorated.
But we still do not know enough to judge the business outcome.
Perhaps the campaign is now producing substantially more conversions and those conversions remain profitable at £42.
Perhaps £42 is unacceptable and the £50 target was simply wrong.
Perhaps lead quality has changed as volume increased.
Perhaps conversions are being measured too early in the customer journey.
Google Ads metrics need commercial context.
A campaign should not be judged purely on whether its cost per conversion is lower than it was last month. It should be judged on whether it is producing worthwhile business outcomes at a cost the business can sustain.
Did Google automatically change your budget or bidding target?
No.
Google says the August rollout does not automatically change an advertiser’s daily budget or Target CPA/Target ROAS.
The targets already configured in the account remain the advertiser’s targets unless somebody changes them.
That is precisely why businesses should review those numbers now.
If a Target CPA was originally set very loosely because the campaign happened to outperform it anyway, the new bidding behaviour can make that forgotten decision more consequential.
How we’d diagnose a sudden August performance change
If a Google Ads campaign has changed noticeably during or after the rollout, we would avoid immediately assuming either that Google is responsible or that the campaign has failed.
We would start by checking:
- Is the campaign Limited by budget?
- Is it using Target CPA or Target ROAS?
- When did the performance change actually begin?
- How did recent actual CPA or ROAS compare with the target configured in the account?
- Were any other meaningful campaign, website or tracking changes made around the same time?
- Has enough time passed to cover one or two normal conversion cycles?
- Has conversion volume changed as well as efficiency?
- Are the conversions producing genuine business value?
That usually gives a much clearer picture than reacting to a single graph.
Don’t blame every August fluctuation on Google
The timing of this update makes it tempting to attribute every unusual result after 17 August to Google’s bidding systems.
That would be a mistake.
Google says the rollout began globally on 17 August 2026, but it is gradual. Not every eligible campaign necessarily changed behaviour at exactly the same moment.
Campaign performance can also change because of competition, search demand, seasonality, conversion tracking, website problems, pricing, offers, budgets, creative, targeting and ordinary statistical variation.
The August update gives advertisers another potential explanation to investigate.
It does not replace proper diagnosis.
The practical takeaway
If a budget-limited Google Ads campaign using Target CPA or Target ROAS has changed during August 2026, Google’s bidding update could genuinely be part of the explanation.
But the most important response is not to start changing everything.
First, check whether the campaign is actually affected.
Then compare the target you told Google to pursue with the performance the campaign had actually been delivering.
If those numbers are significantly different, ask which one reflects the economics of the business.
Give the campaign enough time to cover its normal conversion cycle, be cautious with forecasts during the 17–31 August transition period, and avoid adding unnecessary bidding restrictions purely because the graph moved.
Google Ads is automated, but the commercial target still belongs to the advertiser.
If this update exposes that the target in the account was never really the target the business wanted, correcting that may be more important than trying to reverse every change in performance.
